2.6.0: reiserfs errors: reiserfs_read_locked_inode and friends
Hi, thanks for taking the time to read this.
I have been experimenting between 2.4.23 and 2.6.0-test{5,6,7,8,9,10,11} and now 2.6.0.
Starting with -test11, I noticed often that I could not perform a startx, and as root got permission denied trying to execute or even list some files. I also notice this in 2.6.0. Dropping back to 2.4.23, I can perform the mentioned actions.
My motherboard is an Asus A7M266-D, and I run two AMD XP 1800+ CPU's. I am using a Seagate ST3120023A IDE disk drive.
Here is what I'm seeing in logs:
Dec 18 22:29:40 localhost kernel: blk: queue dfd8fc00, I/O limit 4095Mb (mask 0xffffffff)
Dec 18 22:29:40 localhost kernel: hda: dma_intr: status=0x58 { DriveReady SeekComplete DataRequest }
Dec 18 22:29:40 localhost kernel:
Dec 18 22:29:40 localhost kernel: hda: set_drive_speed_status: status=0x58 { DriveReady SeekComplete DataRequest }
Dec 18 22:29:40 localhost kernel: ide0: Drive 0 didn't accept speed setting. Oh, well.
Dec 18 22:29:40 localhost kernel: blk: queue dfd8f800, I/O limit 4095Mb (mask 0xffffffff)
Dec 18 22:29:40 localhost kernel: is_tree_node: node level 0 does not match to the expected one 2
Dec 18 22:29:40 localhost kernel: vs-5150: search_by_key: invalid format found in block 8397. Fsck?
Dec 18 22:29:40 localhost kernel: vs-13070: reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [373169 373
170 0x0 SD]
Dec 18 22:29:48 localhost kernel: eth0: no IPv6 routers present
Dec 18 22:29:54 localhost kernel: is_tree_node: node level 0 does not match to the expected one 2
Dec 18 22:29:54 localhost kernel: vs-5150: search_by_key: invalid format found in block 8397. Fsck?
Dec 18 22:29:54 localhost kernel: vs-13070: reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [388986 388990 0x0 SD]
Here is what i'm passing to 2.6.0 via grub:
root=/dev/hda6 ide0=autotune ide1=autotune
The kernel claims the reiserfs partitions are "3.6" format.
Any other information I can give to assist?
Hello!
fire-eyes <[email protected]> wrote:
fe> I have been experimenting between 2.4.23 and 2.6.0-test{5,6,7,8,9,10,11} fe> and now 2.6.0.
fe> Starting with -test11, I noticed often that I could not perform a fe> startx, and as root got permission denied trying to execute or even list fe> some files. I also notice this in 2.6.0. Dropping back to 2.4.23, I can fe> perform the mentioned actions.
fe> Dec 18 22:29:40 localhost kernel: hda: set_drive_speed_status:
fe> status=0x58 { DriveReady SeekComplete DataRequest }
fe> Dec 18 22:29:40 localhost kernel: ide0: Drive 0 didn't accept speed fe> setting. Oh, well.
fe> Dec 18 22:29:40 localhost kernel: blk: queue dfd8f800, I/O limit 4095Mb fe> (mask 0xffffffff)
fe> Dec 18 22:29:40 localhost kernel: is_tree_node: node level 0 does not fe> match to the expected one 2
Sounds that somehow some blocks read by reiserfs are zero-filled.
fe> Here is what i'm passing to 2.6.0 via grub:
fe> root=/dev/hda6 ide0=autotune ide1=autotune
What if won't pass "ide0=autotune ide1=autotune" part for 2.6? Does anything changes?
Bye,
Oleg